Direct from Sample Phenotypic Antibiotic Susceptibility Testing _______________________________________ A Thesis Presented to the Faculty of Graduate School at the University of Missouri – Columbia _______________________________________________________ In Partial Fulfillment of the Requirements for the Degree Master of Science _____________________________________________________ by Aiden Lee Dr. Shramik Sengupta, Thesis Supervisor DECEMBER 2018 1 The undersigned, appointed by the dean of the Graduate School, I have examined the Thesis entitled Direct from Sample Phenotypic Antibiotic Susceptibility Testing Presented by Aiden Lee A candidate for the degree of Masters of Science in Biological Engineering and hereby certify that, in their opinion, it is worthy of acceptance. _______________________________________________________________ Dr. Shramik Sengupta, Department of Biological Engineering _______________________________________________________________ Dr. Kevin Gillis, Department of Biological Engineering _______________________________________________________________ Dr. Azlin Mustapha, Department of Food Science 1 Acknowledgements I would like to thank all of the people who have helped me obtain my master’s degree in biological engineering and making this work possible. My biological parents, Jung Rye Choi, Gil Ju Lee, and my brother Yongman Lee have continually supported and encouraged me to better myself both academically and professionally. My best friends Laura Cooper and Nicholas Padilla have also helped me in times of need and providing support. I would like to specially thank Dr. Shramik Sengupta. He has been a great academic mentor and has helped me both personally and professionally for my undergraduate and graduate career. Dr. Sengupta has patiently helped me to learn research, be patient, and learn from my mistakes. As a result, I firmly believe that I have the essential tools to be successful in the professional career. Next, I would like to thank Dr. Kevin Gillis for providing me with an additional guidance in research and giving me an opportunities outside the research. Finally, I would like to thank all the people who worked me together in our lab. Sachidevi Puttaswamy has been providing me with huge support and guidance since I first joined the lab. It was pleasure working with Timothy Butler, Roli Kargupta, and Yongqiang Yang and I learned to appreciate the values of the coworkers. ii Table of Contents Chapter 1. 1. Introduction 8 1.1. Motivation 8 1.2. Proposed Solution 14 2.Materials and Methods 15 2.1 Rationale and overview 15 2.2 Preparation of Bacterial cultures and seeding of the Blood Culture Broth 17 2.3 Finding bacterial loads (CFU/ml) in positive Blood Culture Broth at time of positivity 18 2.4 Microbial isolation using MNPs 18 2.5 Sample preparation for m-EIS study 20 2.6 Choice of bacterial strains, antibiotics and the concentrations tested 21 2.7 Electrical Readings 23 2.7.1 m-EIS 23 2.7.2 Electrical data collection 25 2.7.3 Analysis of the Electrical Data 26 3. Results and discussion: 28 3.1 - CFU counts from positive blood culture samples and Extraction using MNPs: 28 3.2: m-EIS: 30 4. Conclusions and Future Work 41 5.References 43 iii List of Figures Figure 1:Timeline and steps involved in obtaining AST results using current protocols 10 Figure 2: Proposed method for obtaining rapid "Direct from Sample" AST results (MIC values of candidate antibiotics) 14 Figure 3: Overview of Experimental Approach 16 Figure 4:Experimental Approach using MNP and m-EIS 22 Figure 5:Experimental Approach using MNP and m-EIS……………………………. ……27 Figure 6. Data Analysis using the program ZVIEW® 29 Figure 7: MNP extraction of microorganisms 32 Figure 8. Bulk capacitance Changes over time for E.coli 36 Figure 9. Bulk capacitance Changes over time for S.aureus 38 Figure 10. Determination of MIC 42 List of Tables Table 1 Microbial concentrations (CFU/ml) corresponding to measured optical densities (ODs) at selected wavelengths 17 Table 2. Blood Culture Extraction 29 Table 3 38 Table 4 38 iv Abstract Background: To transition from broad-spectrum antibiotic therapy to a more effective “targeted” therapy, i.e. in order to determine whether the infecting strain is resistant or susceptible to individual candidate antibiotics, clinicians and clinical microbiologists need to determine both the pathogen identity (ID) and the Minimum Inhibitory Concentration (MIC) of the candidate antibiotic(s). It is important to note that the pathogen ID or MIC alone is insufficient to yield clinically significant report for patient treatment. Traditionally, tests to determine both the ID and MIC required use of pure bacterial colonies (obtained by plating the patient samples on agar plates) for testing. However, with recent scientific advancement, there are a number of products developed to determine pathogen ID directly from positive patient samples therefore bypassing the need for colony isolation, and delivering results in approximately 6-8 hours after sample collection. However, doing so for the AST remains a challenge, and so improvements in patient outcome have been limited. Methods and Findings: Here, we present a 2-step method to obtain AST-results within few hours of sample collection directly from positive patient blood culture samples (bypassing the need to isolate colonies). The first involves selectively isolating pathogen cells present in samples like blood culture broth using commercially available magnetic-nanoparticles (MNPs) into smaller volumes, thereby bypassing the need for pre-culture. The second involves using microchannel Electrical v Impedance spectroscopy (m-EIS) to monitor multiple aliquots of small volumes (~10μL) of suspension containing bacterial cells, MNPs and candidate-drugs to determine whether cells grow, die, or remain static under the conditions tested. m-EIS yields an estimate for the solution “bulk capacitance” (Cb), a parameter that is proportional to the number of live bacteria in suspension. We are thus able to detect cell death (bactericidal action of the drug) and cell stasis (bacteriostatic action of the drug) in addition to cell-growth. We demonstrate proof-of-principle using E.coli and S.aureus suspended in blood culture broth. Concentration of bacteria extracted from blood culture broth (using MNPs) was set to be ~100,000 CFU /ml with 10 mg/ml of MNPs in growth culture media. These samples were monitored for the presence of bacteriostatic and bactericidal drugs at concentrations below, at, and above known Minimum Inhibitory Concentration (MIC) values. m-EIS data showed data consistent with growth, death, or stasis as expected and/or recorded using plate counts. Electrical signals of growth and death were visible within 3 hours thus allowing us to perform AST testing within 4 hours. Conclusion: We demonstrated “proof of principle” that (a) CFUs counts can be obtained from positive blood culture and (b) extraction can be performed to isolate the microbial cells to increase its concentration, and (c) that the efficacy of candidate drugs on bacteria thus isolated (in suspensions containing MNPs) could be tested in real- time using m-EIS. vi Chapter 1. Introduction 1.1. Motivation For a large number of infectious diseases including, but not limited to bloodstream infections (septicemia),urinary tract infections (UTIs), respiratory infections (including tuberculosis), infected orthopedic implants, etc., patient outcomes are very closely linked to how quickly clinicians can move from broad- spectrum therapy to more effective “targeted” antibiotic therapy. In some cases, quick transition to directed therapy is a life or death issue. For instance: It has been estimated that for a patient in intensive care suffering from severe sepsis (systemic inflammation brought on by bloodstream infections),the chance of survival drops by an average of ~7% for every hour of delay in switching from broad-spectrum antibiotics (empiric therapy) to an effective antibiotic (directed therapy).[21] In order to deliver effective antibiotics, clinicians need to know both the identity of the pathogen causing the infection and the Minimum Inhibitory Concentration (MIC) of candidate antibiotics against the pathogens to effectively transition from broad-spectrum to targeted antibiotic therapy. The traditional protocol to obtain both pathogen ID and MIC using instruments like the Vitek™ (Biomerieux) and Phoenix™ (Becton Dickinson) is shown in Figure 1(B). Both ID and AST (MIC determination) involve physically observing bacteria growing (or failing to grow) in specially prepared media: those with different carbon source / nutrients profile for pathogen ID, and when challenged with candidate antibiotics at various concentrations for AST. 1 (A) (B) Figure 1. Timeline and steps involved in obtaining AST results using current protocols (A) (TOP) Time-line of steps needed to obtain phenotypic Antibiotic Susceptibility Testing (AST results) i.e. values of the Minimum Inhibitory Concentration (MIC) of candidate drugs for a bloodstream infection for three cases. Case 1: using traditional protocol (culture using BACTEC TM / BacT-Alert TM etc., followed by ID and AST using Phoenix TM / VITEK TM, etc.); Case 2: Upon using emerging / recently approved molecular methods for microbial ID (Biofire TM, E-plex TM etc.) after traditional blood culture, while using with traditional methods for AST; Case 3: Upon using proposed “direct from sample” method for AST along with molecular methods for ID. (B) (Bottom): Steps involved in obtaining